Gujarati arabic typing rules

ISO/IEC JTC1/SC2/WG2 N4574L2/14-1312014-05-02

Title: Proposal to Encode Gujarati Signs for the Transliteration of Arabic in ISO/IEC 10646Source: Script Encoding Initiative (SEI)Author: Anshuman Pandey ([email protected])Status: Liaison ContributionAction: For consideration by UTC and WG2Date: 2014-05-02

1 Introduction

This is a proposal to encode six additional characters in the ‘Gujarati’ block of the Universal Character Set(ISO/IEC 10646):

◌ 0AFA GUJARATI SIGN SUKUN

◌ 0AFB GUJARATI SIGN SHADDA

◌ 0AFC GUJARATI SIGN MADDAH

◌ 0AFD GUJARATI SIGN THREE-DOT NUKTA ABOVE

◌ 0AFE GUJARATI SIGN CIRCLE NUKTA ABOVE

◌ 0AFF GUJARATI SIGN TWO-CIRCLE NUKTA ABOVE

The characters are proposed for contiguous allocation in the six empty code points at the end of the ‘Gujarati’block after the approved, but not yet encoded characterૹ +0AF9 . The location issuitable because , like the six characters proposed here, is used for transliteration.

2 Description

These signs are used for the transliteration of the Arabic script into Gujarati by Ismaili Khoja communities.They are used for representing Arabic letters and signs for which correspondences do not exist in Gujarati.They were devised in the late 19th century and are standard elements of the Gujarati orthography used by theIthnashari Khoja (“Twelver Shia”) and the Agakhani Khoja communities. The creation of the full set and thefirst documented printing of these signs was undertaken by the Ithnashari Khoja publisher Gulāmalī Ismāʾilof Bhavnagar, Gujarat in 1901. The signs are used in manuscripts and in printed materials, predominantlyin Khoja texts such as the “Agakhani Dua” and Gujarati-script versions of the Qurʾān.

1

Proposal to Encode Gujarati Signs for the Transliteration of Arabic in ISO/IEC 10646 Anshuman Pandey

The ◌ , ◌ , and ◌ occur with severalbase letters; however, the distribution of the three nukta signs is limited to particular base letters. In theavailable Ithnashari sources the ◌ - is used with several base letters,but ◌ occurs with only two consonants and ◌ - occurs with only one letter. There are two approaches to encoding these latter twonukta signs: 1) as combining signs, or 2) as atomic characters consisting of each sign combined with arespective base letter. This proposal recommends the first option in order to provide flexibility in the usageof these signs.

The ◌ - is used in Ithnashari orthography only in combinationwith ઝ , in the form ઝ , for the transliteration of ظ +0638 ,the pharyngealized voiced dental fricative [ðˤ]. Given this limited usage it may be practical to encode ઝ asan atomic character, ie. * . The situation is identical to that of ૹ , which was proposed by Vinodh Rajan in May 2013 and approved by the Unicode TechnicalCommittee (UTC) for future encoding at +0AF9 in the Gujarati block (see L2/13-143). The letter ૹ isused for transliterating �� +10B32 . In principle, ૹ is aprecomposed character consisting of the base જ +0A9C combined with the sign ◌,which is a three dot variation of ◌ +0ABC . Ideally, the UTCwould have consideredencoding this ◌ three-dot nukta as a combining mark instead of encoding ૹ as anatomic character. This would have made it possible to use ◌with other base letters, if needed. Indeed, Rajanhad initially proposed the encoding of ◌ as * so that ૹ might be representedas sequence of a base letter and combining mark (see L2/13-066). But, it seems that ૹ was encoded as theatomic on account of its correspondence to ॹ +0979 ,which is also used for transliterating Avestan �� . The encoding of ૹ as an atomic letter, however,does not account for the possibility that ◌ may occur in some source with another letter, eg. ઝ. Such a casewould necessitate either the encoding of that combination as a separate atomic character or, ultimately, theencoding of ◌ as a combining mark. For this reason, although ◌ - occurs in Ithnashari orthography with only ઝ , it is recommended that it, along with◌ , be encoded as a combining sign instead of as atomic compositionswith base letters in order to account for the possibility that they may be used with other base letters in othersources.

3 Characters Proposed

3.1 GUJARATI SIGN SUKUN

The ◌ is used for indicating a pause during recitation. It can occur with vowel lettersand consonants. When it occurs with a consonant, it can also represent a vowel-less consonant. Whenused for the latter, it behaves similar to ◌ +0ACD in its function of silencing theinherent vowel of a consonant, but it does not possess the control properties of . The Gujarati sukuncorresponds to +0652 . It is used in encoded text as follows:

અ <અ , ◌ sukun>

સ <સ , ◌ sukun>

The sukun occurs, for example, in the phrase ولكن / વલા કન , highlighted in dark green in figures 1 and 2.

2


3.2 GUJARATI SIGN SHADDA

The ◌ represents consonant gemination. It is modelled upon +0651 and corresponds to ◌ 𑈷 +11237 . It can also occur in conjuncts above half-letters.

સ <સ , ◌ shadda>

સ <સ , ◌ shadda, ◌ , સ , ◌ shadda>

The shadda occurs, for example, in the phrase ان / ઇન, highlighted in cyan in figures 1 and 2.

3.3 GUJARATI SIGN MADDAH

The ◌ represents the lengthening or sustain of a vowel during recitation. It ismodelled upon the +0653 . It is used in encoded text as follows:

ક <ક , ◌ maddah>

કા <ક , ◌ા , ◌ maddah>

ક <ક , ◌ , ◌ maddah>

The maddah occurs, for example, in the word اولئك /ઓલાએક, highlighted in red in figures 1 and 2.

3.4 GUJARATI SIGN THREE-DOT NUKTA ABOVE

The ◌ - is used for representing the Arabic letters shown below. Itcan occur with vowel and consonant letters. The sign corresponds to ◌𑈶 +11236 .

Arabic Gujarati Encoded sequence

ع અ <અ , ◌ three-dot nukta above>

ع ઇ <ઇ , ◌ three-dot nukta above>

ق ક <ક , ◌ three-dot nukta above>

خ ખ <ખ , ◌ three-dot nukta above>

غ ગ <ગ , ◌ three-dot nukta above>

ط ત <ત , ◌ three-dot nukta above>

ص સ <સ , ◌ three-dot nukta above>

ض ઝ <ઝ , ◌ three-dot nukta above>

The three-dot nukta occurs, for example, in the word مرض / મરઝન , highlighted in purple in figures 1 and 2.

3


When both three-dot nukta above and a dependent vowel sign occur with a base letter, the three-dot nuktaabove occurs before the vowel sign. The rationale for the encoding order is that, identical to the regular , the three-dot nukta above is a consonant modifier and its usage affects the value ofthe consonant independently of any accompanying vowel sign.

ક <ક , ◌ three-dot nukta above, ◌ >

3.5 GUJARATI SIGN CIRCLE NUKTA ABOVE

The ◌ is used for representing the following two Arabic letters:


ث સ <સ , ◌ circle nukta above>

ذ ઝ <ઝ , ◌ circle nukta above>

The circle nukta above occurs, for example, in the phrase اذا و اذا) (و / વએઝા , highlighted in blue in figures 1and 2.

3.6 GUJARATI SIGN TWO-CIRCLE NUKTA ABOVE

The ◌ - is used for representing the following Arabic letter:


ظ ઝ <ઝ , ◌ two-circle nukta above>

The two-circle nukta above occurs, for example, in the word عظيم / અ ઞી મ, highlighted in light green infigures 1 and 2.

4 Considerations for Rendering

More than one of the proposed marks may occur with a single base letter. In such cases it is necessary toadjust the placement of the signs in order to prevent clashing. Generally, the signs are placed horizontally.

Ordering When sukun occur together with one of the three above-base nukta signs or with shadda, thelatter occurs first in the encoded representation, but it is positioned to the right of the sukun in the output:

અ <અ , ◌ three-dot nukta above, ◌ sukun>

ઝ <ઝ , ◌ two-circle nukta above, ◌ sukun>

સ <સ , ◌ shadda, ◌ sukun>

The same principle applies to the co-occurrence of shadda and one of the three above-base nukta signs:

4


અ <અ , ◌ three-dot nukta above, ◌ shadda>

ઝ <ઝ , ◌ two-circle nukta above, ◌ shadda>

Glyph Adjustment When ◌ three-dot nukta above occurs with vowel signs that extend above and overthe body of the consonant — ◌ +0AC7 , ◌ +0AC8 ,◌ો +0ACB , ◌ૌ +0ACC — then it is reduced in sizeand rotated in order to prevent clashing and to accommodate fit. The same size reduction takes places whenthe - occurs with ◌ and the width of the two signs exceeds the width of the basecharacter:

સ → સ

<સ , ◌ three-dot nukta above, ◌ >

હ → હ <હ , ◌ three-dot nukta above, ◌ sukun>

5 Characters Not Proposed

The sources show a mark resembling, which is used as a sectioning mark. This mark could be a glyphicvariant of ٭ +066D . It corresponds to the +06DD used in the parallel Arabic. The occurs in three different contexts: 1) independently, 2) followed by oneor more digits (૩૭), and 3) with a superscript consonant letter. The source shows two superscript letters,eg. અ and મ , written above the mark:

અand

મ.

The section mark is not proposed for encoding as a separate character at present. The ٭ +066D may be used, with appropriate changes to the glyph. The superscript letters may berepresented as rich text. However, section mark may be proposed for encoding in the future if additionalresearch shows it to be a distinct character used in Gujarati transliteration of Arabic.

6 Character Data

Character Properties Character properties given in the data format of UnicodeData.txt:

0AFx;GUJARATI SIGN SUKUN;Mn;0;NSM;;;;;N;;;;;0AFx;GUJARATI SIGN SHADDA;Mn;0;NSM;;;;;N;;;;;0AFx;GUJARATI SIGN MADDAH;Mn;0;NSM;;;;;N;;;;;0AFx;GUJARATI SIGN THREE-DOT NUKTA ABOVE;Mn;7;NSM;;;;;N;;;;;0AFx;GUJARATI SIGN CIRCLE NUKTA ABOVE;Mn;7;NSM;;;;;N;;;;;0AFx;GUJARATI SIGN TWO-CIRCLE NUKTA ABOVE;Mn;7;NSM;;;;;N;;;;;

Linebreaking Properties Linebreaking properties given in the data format of LineBreak.txt:

0AFx;CM # GUJARATI SIGN SUKUN0AFx;CM # GUJARATI SIGN SHADDA0AFx;CM # GUJARATI SIGN MADDAH0AFx;CM # GUJARATI SIGN THREE-DOT NUKTA ABOVE0AFx;CM # GUJARATI SIGN CIRCLE NUKTA ABOVE0AFx;CM # GUJARATI SIGN TWO-CIRCLE NUKTA ABOVE

Syllabic Categories Syllabic categories given in the data format of IndicSyllabicCategory.txt:

5


# Indic_Syllabic_Category=Nukta0AFx ; Nukta # Mn GUJARATI SIGN THREE-DOT NUKTA ABOVE0AFx ; Nukta # Mn GUJARATI SIGN CIRCLE NUKTA ABOVE0AFx ; Nukta # Mn GUJARATI SIGN TWO-CIRCLE NUKTA ABOVE

# Indic_Syllabic_Category=Gemination_Mark0AFx ; Gemination_Mark # Mn GUJARATI SIGN SHADDA

Matra Categories Matra categories given in the data format of IndicMatraCategory.txt:

# Indic_Matra_Category=Top0AFx..0AFx ; Top # Mn [6] SIGN SUKUN .. SIGN TWO-CIRCLE NUKTA ABOVE

7 References

Ismā’il, Gulāmalī. 1901–1903. Anvārūl bayān phī taphsīril kura’ān. 3 vols. Bhāvnagar: Isnā’aśarī ilēkṭrīkprīnṭīng prēs.

Rajan, Vinodh. 2013a. “Proposal to Encode Gujarati Sign Triple Nukta”. L2/13-066. http://www.

unicode.org/L2/L2013/13066-gujarati-triple-nukta.pdf

———. 2013b. “Proposal to Encode Gujarati Letter ZHA”. L2/13-143. http://www.unicode.org/L2/

L2013/13143-gujarati-zha.pdf

8 Acknowledgments

I am grateful to Iqbal Akhtar (Florida International University, Miami) for bringing these Gujarati signs tomy attention and for providing specimens of usage. I would also like to thank Roozbeh Pournader (Google)for offering some insights into the Arabic orthography used in the Qurʾān.

This project was made possible in part with support from the Script Encoding Initiative at the University ofCalifornia, Berkeley.

6


Figure 1: Page of the Qurʾān in Gujarati showing usage of the Arabic transliteration marks (fromGulāmalī Ismāʾil 1901: 36). Arabic parallel shown in figure 2.

7


Figure 2: Page of the Qurʾān in Arabic (from Gulāmalī Ismāʾil 1901: 35). Gujarati parallel shownin figure 2.

8

Documents

Gujarati arabic typing rules